UDC

004.932

Authors

Zykov Aleksey Nikolaevich
Shiprepair Centre Zvyozdochka JSC
Mashinostroiteley ave., 12, Severodvinsk, Arkhangelsk Region, 164509, Russian Federation;
e-mail: alexzikov@gmail.com

Abstract

Fractal coding is an efficient method of image compression based on the representation of the image by contractive affine transformations in the image space, where the fixed point is close to the original image. Fractal coding of images is based on the self-similar and self-affine sets and uses the internal redundancy between the self-similar parts of the image. The encoding process is the use of the system of the iterated piecewise-defined functions (PIFS – Partitioned Iterated Function Systems) in the construction of the operator, which will represent the image of the encoded image. The key problem of the existing methods of fractal image coding is a long processing time, as the finding of a similarity mapping between the domain and range areas is a rather complex task, requiring significant computing resources. The most effective method is to make a classification of the image areas, and then to find a similarity of mapping. Classification of the areas allows us to transfer a global search into a fractional search. The finding of a similarity mapping is performed only among the blocks belonging to the same class that causes speeding up time of encoding significantly. The key point is that the classification criteria should describe in the best way the information of the whole image and each area of the image, as well as the degree of similarity of the image areas. The classification function includes the rules determining the number of classes and classification accuracy. The new method proposes to use the operator of the local binary patterns (Local Binary Pattern – LBP) in the problem of fractal coding of the gray-scale images. LBP converts the image into an array of binary codes describing the pixel neighborhood. The local binary patterns are based on the local binary model, offer an effective way to analyze the image texture and are considered as its effective characteristic. The experiments demonstrate the relevance of this method, which speeds up the process and improves the quality of the reconstructed image.

Keywords

image compression, fractal coding, image analysis, local binary patterns

References

1. Barnsley M., Hurd L. Fractal Image Compression. Wellesley, 1993. 
2. Jacquin A.E. Fractal Image Coding: A Review. Proceedings of the IEEE, 1993, vol. 81(10), pp. 1451–1465. 
3. Welstead S. Fractal and Wavelet Image Compression Techniques. Tutorial Texts in Optical Engineering, 1999, vol. 40. 
4. Ojala T., Pietikainen M., Maenpaa T. Multi Resolution Gray-Scale and Rotation Invariant Texture Classification with Local Binary Patterns. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2002, vol. 24, no. 7, pp. 971–987. 
5. Ahonen T., Hadid A., Pietikainen M. Face Recognition with Local Binary Patterns. Proc. European Conf. Computer Vision, 2004, pp. 469–481. 
6. Takala V., Ahonen T., Pietikainen M. Block-Based Methods for Image Retrieval Using Local Binary Patterns. Scandinavian Conf. on Image Analysis. Finland, Joensuu, 2005, pp. 882–891.